Dreamforce 2026: Dates, Highlights & Everything Worth Knowing

Dreamforce 2026: Dates, Venue and How to Watch

Dreamforce 2026 recap: dates, venue, AIforce, Koa, seven new Agentforce agents, Google Cloud and AWS partnerships, and what it all means for your business.

Dreamforce is the one week each year when Salesforce shows its hand, and Dreamforce 2026 may be the most consequential edition since Agentforce was introduced. The headline is simple: Salesforce no longer wants to be the screen you open. It wants to be the trusted business context that powers AI wherever your teams already work.

At Avanciers, our consultants have been tracking every keynote, press release and product session so that our clients do not have to. Here is a clear, business-first summary of Dreamforce 2026: when it happened, what was announced, and what you should do about it.

Dreamforce 2026: Dates, Venue and How to Watch

• Dates: September 15–17, 2026
• Venue: Moscone Center, San Francisco
• Theme: The Agentic Enterprise – humans and AI agents working together across data, applications and processes
• Scale: More than 1,400 sessions, with the opening keynote delivered by Chair and CEO Marc Benioff on September 15
• Virtual access: The free Salesforce+ stream carried the main keynotes, 400+ sessions and live launch coverage for Agentforce, Customer 360, Data 360 and Slack, and on-demand replays remain available after the event

The sponsor roster was telling. PwC, Boston Consulting Group, KPMG, IBM, AWS and Slalom all featured prominently, a signal that the global services ecosystem now treats agentic AI readiness as a core practice rather than an experiment.

The Big Idea: The Enterprise AI Harness

Before a single product was named on stage, Salesforce published the architecture that ties this year together. Announced on September 10, the Enterprise AI Harness is a composable framework spanning six capabilities – context, agency, action, governance, security and models so that organizations do not have to rebuild those foundations for every new agent or AI interface.

Understanding this framework makes every subsequent announcement easier to place. Each headline product below is essentially one of those six layers given a name.

Top Announcements from Dreamforce 2026

1. AIforce: AI Replaces the UI

The flagship reveal. AIforce is a live interface layer that brings Salesforce data, workflows, business logic, permissions and governance to any AI surface. It launched with three entry points:

• Claudeforce, born from the Salesforce–Anthropic partnership announced in August
• Slackforce, positioning Slack as the conversational front door to the enterprise
• Agentforce Coworker, an AI teammate that works alongside employees

Two design choices stand out for enterprise buyers. Every request runs on the user’s existing permissions, so agents see only what that person can see. And Salesforce has committed to zero data retention, meaning business data is used to answer a question but not retained by the model provider.

2. Koa: Salesforce’s First CRM Reasoning Model

Built with NVIDIA on the Nemotron model family, Koa is a reasoning model designed specifically for multistep CRM tasks. Salesforce says it was trained on 27 years of CRM intelligence using synthetic scenarios, with no customer data used in training, and that it runs inside Salesforce’s own infrastructure. For regulated industries, that training and inference boundary is a meaningful differentiator.

3. Seven Job-Ready Agentforce Agents

Announced days before the keynote, Salesforce introduced a portfolio of pre-built agents mapped to real job roles:

• Casey – customer service across voice, SMS, WhatsApp and chat
• Paige – IT and HR requests
• Carter – e-commerce
• Marshall – supply chain and back-office orchestration
• Piper – inbound sales qualification
• Fin – complex customer experience workflows
• Hunter – outbound sales (in pilot, with general availability slated for November 2026)

Six of the seven are generally available today. For businesses that have struggled to define a first agent use case, these give a concrete, measurable starting point.

4. Headless 360 and Agent Fabric

Announced on August 19, Headless 360 turns every Salesforce cloud into reusable capabilities that authorized AI agents running in Agentforce, Claude, ChatGPT, Cursor and other platforms can discover and invoke through open standards such as MCP. Agent Fabric, MuleSoft’s control plane, provides governance and identity across multi-vendor agents from Amazon Bedrock, Google Vertex AI and Microsoft Copilot Studio.

Together they dramatically widen the surface through which AI can touch your Salesforce org. That is powerful, and it makes permission hygiene more important than ever.

5. Expanded Google Cloud and AWS Partnerships

On the same day, Salesforce deepened ties with both hyperscalers. The Google Cloud deal connects Agentforce and Gemini Enterprise over MCP and brings Salesforce workloads to Google Cloud via Hyperforce. The AWS expansion puts Salesforce context inside Amazon Quick, brings AWS agents into Slack, adds model choice through Bedrock, and links Agentforce Voice with Amazon Connect. Expanded Data 360 Zero Copy functionality is scheduled for fall 2026.

The message is deliberate neutrality: choose your cloud, choose your model, and keep Salesforce as the governed context layer underneath.

6. Customer Proof and Community Commitments

Adecco Group announced an Agentforce Coworker rollout across more than 40 countries following pilots in the UK and France—a useful template for phased adoption. Salesforce also reported that Agentforce has reached 30,000 customers two years after launch, and pledged $27 million to education initiatives.

What Dreamforce 2026 Means for Your Business

Strip away the product names and three practical implications remain.

Your permission model is now your AI security boundary. AIforce and Headless 360 both rely on existing sharing rules. Years of profile clones and “temporary” access grants will be exercised by far more requests than they were designed for. An access audit is the single most valuable pre-AI investment you can make.

Data quality determines AI quality. Every announcement this week assumes a unified, trusted data foundation in Data 360. Agents reasoning over conflicting records will confidently produce the wrong answer. If you have not yet unified customer data, that is step one.

Start narrow, measure and scale. The job-ready agents and the Adecco rollout point the same way: pick one queue you already measure, deploy one agent with clear guardrails, prove value, then expand.

How Avanciers Can Help

Dreamforce sets the vision; implementation determines the return. Our Salesforce consulting team helps organizations translate this year’s announcements into a practical roadmap:

• Agentic readiness assessment covering data, permissions and governance
• Data 360 unification to give agents trusted, real-time context
• Agentforce pilot delivery using the new job-ready agents where they fit your workflows
• Integration architecture across MuleSoft, Google Cloud and AWS
• Managed services to keep your org healthy as new capabilities land

Want a Dreamforce 2026 briefing tailored to your Salesforce environment? Schedule a consultation with our experts and turn this year’s announcements into next quarter’s results.